Quaternary Science Reviews 234 (2020) 106248 Contents lists available at ScienceDirect Quaternary Science Reviews journal homepage: www.elsevier.com/locate/quascirev Timing of formation of neoglacial landforms in the South Shetland Islands (Antarctic Peninsula): Regional and global implications * David Palacios a, , Jesus Ruiz-Fernandez b, Marc Oliva c, Nuria Andres a, Jose M. Fernandez-Fern andez a, Irene Schimmelpfennig d,Laetitia€ Leanni d, Benjamín Gonzalez-Díaz b, ASTER Teamd, e a Department of Geography, Complutense University of Madrid, Spain b Department of Geography, University of Oviedo, Spain c Department of Geography, Universitat de Barcelona, Spain d Aix Marseille Univ, CNRS, IRD, INRAE, Coll France, CEREGE, Aix-en-Provence, France e Consortium: Georges Aumaître, Didier Bourles, Karim Keddadouche, France article info abstract Article history: The timing of neoglacial advances in the Antarctic Peninsula (AP) is not yet well constrained. Accurate Received 18 November 2019 temporal reconstruction of Neoglaciation in the AP is needed to better understand past glacial responses Received in revised form and regional and global teleconnections during the Holocene. Here, we examine all available information 18 February 2020 about neoglacial advances in the South Shetland Islands (SSI) as well as in the broader geographical Accepted 26 February 2020 context of the AP region and Antarctic continent. In order to shed light on the contrasting chronologies Available online xxx existing for neoglacial advances in these regions, we focused on a case study where a detailed picture of the Holocene deglaciation was already available. Lake sediments revealed that Byers Peninsula, west of Keywords: Antarctic peninsula Livingston Island (SSI), was fully deglaciated during the Holocene Thermal Maximum. To complement fi Byers peninsula this approach, we identi ed glacially polished bedrock surfaces, erratic boulders and a moraine ridge Neoglaciation near the present front of the glacier in the SE corner. We applied cosmogenic ray exposure (CRE) dating Surface exposure dating using in situ 36Cl for basalt rocks and 10Be for granitic rocks in: (i) 8 samples from glacial erratic and ice- rafted boulders, (ii) 2 samples from moraine boulders, (iii) 2 samples from polished bedrock surfaces, and (iv) 1 sample from an erratic boulder deposited on one of these surfaces. The CRE dates indicate that the onset of deglaciation started around 9.9 ± 1.2 ka, with two phases of glacier expansion during the Mid- Late Holocene forming moraines at ~4.1 ± 0.5 and ~1.0 ± 0.2 ka, respectively. The main neoglacial ad- vances in the AP and the SSI were mostly synchronous and coincided with cold periods, as shown by other records (e.g. glacio-isostatic marine terraces, marine and lake sediments). In addition, these periods of glacial expansion show a similar timing to those recorded in the Arctic. These results suggest that Neoglaciation was driven by global climate forcing in both polar areas despite temporal variations at regional and local scale. © 2020 Published by Elsevier Ltd. 1. Introduction AP region than the Earth’s average (Vaughan et al., 2003; Turner et al., 2005). This dramatic temperature rise caused fast melting The Antarctic Peninsula (AP; Fig. 1) is affected by intense of glaciers and a generalized retreat of glacier fronts (Vaughan and warming, which is recorded through reliable meteorological mea- Doake, 1996; Cook et al., 2005; Kunz et al., 2012; Pritchard et al., surements since approximately 1950. This most recent period has 2012). However, since 1998 a tendency towards cooling in the AP been called Recent Rapid Warming (RRW), and is characterized by and a return to positive mass glacial balances has been detected an accelerated temperature increase being four times greater in the (Navarro et al., 2013; Engel et al., 2018). This shift is associated with an increase in the frequency of low pressure systems that has led higher precipitation and lower summer temperatures in the region, particularly in the N and NE AP region, including the South Shet- * Corresponding author. land Islands (SSI) (Turner et al., 2016; Sancho et al., 2017; Oliva E-mail address: [email protected] (D. Palacios). https://doi.org/10.1016/j.quascirev.2020.106248 0277-3791/© 2020 Published by Elsevier Ltd. 2 D. Palacios et al. / Quaternary Science Reviews 234 (2020) 106248 Fig. 1. Location of areas cited in the text related to Antarctica. A) Antarctic Ice Sheet; B) Ross Ice Shelf; C) Antarctic Peninsula; D) Graham Land. This figure is available in colour in the online version. D. Palacios et al. / Quaternary Science Reviews 234 (2020) 106248 3 et al., 2017a)(Fig. 2). Therefore, the question arises whether RRW is the Neoglaciation chronology in the AP and compared it with the the beginning of a trend towards deglaciation in the AP, or a brief neoglacial advances in Patagonia. warming within the long-term cooling trend affecting this region The Byers Peninsula, on the western end of Livingston Island, is over the last millennia, continuation of the long-term cooling the largest deglaciated terrestrial area in the SSI (Fig. 2). In this area, started since the Neoglaciation (Vaughan et al., 2003; Bentley et al., lake records suggest that the retreat of the Rotch Dome Glacier 2009). However, the timing, distribution, evolution and origin of occurred throughout the Holocene (Toro et al., 2013; Oliva et al., the neoglacial advances in the AP region is still poorly understood 2016; Ruiz-Fernandez and Oliva, 2016), although the existence of because of: (i) the lack of historical information on this continent, neoglacial landforms, such as moraines, is not yet evidenced. The (ii) the limited number of terrestrial natural archives as sources of application of cosmogenic radiation exposure (CRE) methods e that palaeoenvironmental information, and (iii) the difficulties of dating have not been applied yet to neoglacial landforms in the SSI e offers these glacial advances in Antarctica (Davies et al., 2012). new possibilities to obtain a detailed chronology for the develop- Porter and Denton (1967) proposed the generalization of the ment of these neoglacial landforms. The knowledge of the age of term Neoglaciation to refer to the glacial advances that occurred moraines formed in neoglacial advances in the Byers Peninsula can after the Holocene Thermal Maximum (HTM: 11-5 ka, Renssen provide important information on the possible synchrony of these et al., 2009) until the end of the Little Ice Age (LIA). Evidence of advances within the context of the AP. Late Holocene glacial advances has been found in a great variety of The aim of this paper is to map and date neoglacial landforms in latitudes and continents (Solomina et al., 2015), widespread also in the SSI as well as explore their paleoclimatic implications in the mountains of the Southern Hemisphere (Clapperton and Sugden, context of the Antarctic continent. Firstly, we review the current 1988; Porter, 2000) and Antarctica (Clapperton and Sugden, 1988; state of knowledge of Neoglaciation in Antarctica, focusing mainly Clapperton et al., 1989; Ingolfsson et al., 1998). However, delimiting in the AP. In order to provide the most detailed picture on the pa- the end of the HTM and the onset of the neoglacial chronology in leoclimatic evolution in this region, we examined all available re- the Polar Regions is a matter of great complexity (Kaufman et al., cords from ice cores, deep marine sediment cores, lake sediments 2004; Renssen et al., 2009, 2012). Similarly, the concept of Neo- and glacial landforms. Subsequently, we focused on an area of the glaciation in the Arctic has been proposed as a tendency to climate AP region where information about Holocene glacial advances is cooling (McKay et al., 2018). However, phases of glacial advances in still absent: the Byers Peninsula. In this peninsula, we have mapped the Arctic during Neoglaciation alternated with warm periods and the spatial distribution of possible neoglacial moraines, analyzed glacial retreat, occurring asynchronously across the region without the geomorphological setting, and applied CRE dating to these a clear regional pattern (McKay et al., 2018). landforms. As in the Arctic, the chronology of Neoglaciation varied signifi- cantly within the AP and throughout the Antarctic continent (Bentley and Hodgson, 2009). According to the available glacio- 2. The Neoglaciation in the Antarctic Peninsula and South marine sedimentary records, cold phases during Neoglaciation in Shetland Islands the AP region were short, not exceeding 0.5 ka (Yoon et al., 2010), and alternated with longer warm periods (Davies et al., 2012). The A recent synthesis of the evolution of the West and East Ant- first synthesis on climatic and glacial evolution in the AP during the arctic Ice Sheets shows that glacial mass loss was intense from 9 Holocene was carried out by Ingolfsson et al. (1998) who suggested until 7-6 ka in both areas (Mackintosh et al., 2011, 2014), mainly due the first neoglacial advances after 5 ka. Ingolfsson et al. (2003) to ocean warming, although the long-term glacial shrinking had proposed several other neoglacial advances between 3 and 1 cal already begun before 10 ka (Larter et al., 2014). (Fig. 1, Table 1). A ka BP. Bentley et al. (2009, 2014) reported a generalized warm wide range of climatic proxies provides data on the occurrence of period in the AP between 4.5 and 2.8 cal ka BP followed by a various cold periods in the AP for the last millennia. relatively widespread cold period between 2.5 and 1.2 cal ka BP. In order to constrain the magnitude and the chronology of Bentley and Hodgson (2009) highlighted that Late Holocene cold neoglacial advances and retreats in the SSI, we first review the phases in different sites across the AP were not synchronous, even climatic evolution over the Mid- and Late Holocene in the AP region in close regions. Hall (2009) provided a synthesis of Holocene based on different environmental sources: glacial evolution for the entire Antarctic continent, including the AP, with many references to neoglacial landforms.